Worries mount about another state AI law preemption
Just_Super via Getty Images
By
Chris Teale
Another federal bill — this time in the Great American AI Act — has state lawmakers on edge that Congress will once again try to take away their powers to legislate.
Congress has put forward another effort to preempt states’ artificial intelligence laws, leaving leaders once again on edge and fearful they will be prevented from regulating the technology and mitigating its worst effects.
Reps. Jay Obernolte and Lori Trahan introduced a draft measure earlier this month as part of their Great American AI Act that would allow the federal government to preempt state AI laws for three years.
It’s the latest attempt to block states from regulating AI, both from the administration and Congress. And opponents said it, like every other attempt to tie state legislatures’ hands, cannot be allowed to proceed.
“The current draft of the bill preempts and forbids states from passing laws that regulate the development of AI,” Zephyr Teachout, a professor of law at the Fordham Law School and a former candidate for statewide offices in New York, said during a press conference hosted by the nonprofit Americans for Responsible Innovation. “This is an incredibly broad preemption regime.”
Congress and President Donald Trump’s administration have tried before to preempt states from regulating AI, whether it be through various versions of the One Big Beautiful Bill Act that were eventually quashed, or via an executive order issued last year that was threatened with legal action. The Federal Communications Commission has also said it could get involved in preempting states’ AI laws.
Meanwhile, in the absence of concrete federal action, states have pushed ahead to legislate on AI themselves, especially in areas like child safety, elections, worker safety and surveillance, among others. Any federal AI standard must prioritize keeping people safe and protecting their civil rights, speakers at the press conference said This bill does not, and states have been trying to do so.
“Leaving civil rights behind is unacceptable,” said Alejandra Montoya-Boyer, vice president of the Center for Civil Rights & Technology at The Leadership Conference on Civil and Human Rights. “That omission means that people across the nation will be at risk for data-driven discrimination. People will get locked out of quality jobs, housing, educational opportunities and health care, or can even be wrongfully jailed or imprisoned because of faulty AI systems that supercharge discrimination at massive scale and scope.”
That opposition escalated this week, as ARI released an open letter from more than 200 bipartisan lawmakers from 42 states urging Congress to oppose any preemption of state AI laws. The signatories noted that state legislators are on the front lines of protecting their residents from the technology’s harms and should be allowed to continue that work.
“The Great American AI Act’s preemption provision promises to override state AI laws related to the development of any AI model without implementing a federal framework to replace current state protections,” the coalition of state lawmakers wrote in the letter. “Such a sweeping category of laws would include measures addressing AI models trained on copyrighted works or child abuse content, legislation ensuring that AI is not developed in a way that discriminates against workers or homebuyers, and state bills that protect the privacy of consumers. Not only is the provision extremely broad as written, but the tech industry will almost certainly weaponize such a provision in court to strike down state measures not intended to fall within the scope of GAAIA.”
The open letter from legislators pointed out previous mistakes made by failing to regulate technology and said those mistakes must not be repeated.
“For years, state lawmakers have worked tirelessly to address the risks and harms of social media created by Big Tech companies,” the letter says. “The lessons of the social media era are clear: allowing Silicon Valley to write its own rulebook leaves industry unaccountable and leaves American families vulnerable to AI’s dangers.”
Other opponents of preemption agreed that state lawmakers remain the best people to regulate, especially as they see the effects on their constituents more quickly than their federal counterparts.
“The preemption is what makes this bill illegitimate as a democratic matter, as we approach the 250th anniversary of the Declaration of Independence,” said Teachout during the press conference. “We have to return to the core principle of people governing themselves, and that means people being able to respond in real time to the most significant technological development in generations.”